See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273124327

Sediment bypass tunnels to mitigate reservoir sedimentation and restore

sediment continuity

Chapter · September 2014

DOI: 10.1201/b17397-27

CITATIONS READS

14 586

4 authors:

Robert Michael Boes Christian Auel

ETH Zurich ILF Consulting Engineers, Innsbruck, Austria
118 PUBLICATIONS   695 CITATIONS    31 PUBLICATIONS   135 CITATIONS   

SEE PROFILE SEE PROFILE

Michelle Müller-Hagmann Ismail Albayrak

ETH Zurich ETH Zurich
15 PUBLICATIONS   43 CITATIONS    60 PUBLICATIONS   386 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Sediment transport measurements with geophone sensors View project

FLEXSTOR - Solutions for flexible operation of storage hydropower in changing environmental and market conditions View project

All content following this page was uploaded by Christian Auel on 22 April 2015.

The user has requested enhancement of the downloaded file.

 Reservoir Sedimentation – Schleiss et al. (Eds)
© 2014 Taylor & Francis Group, London, ISBN 978-1-138-02675-9

Sediment bypass tunnels to mitigate reservoir sedimentation

and restore sediment continuity

R.M. Boes, C. Auel, M. Hagmann & I. Albayrak

Laboratory of Hydraulics, Hydrology and Glaciology (VAW), ETH Zurich, Switzerland

ABSTRACT:  Worldwide, a large number of reservoirs impounded by dams are rapidly

filling up with sediments. As on a global level the loss of reservoir volume due to sedimenta-
tion increases faster than the creation of new storage volume, the sustainability of reservoirs
may be questioned if no countermeasures are taken. This paper gives an overview of the
amount and the processes of reservoir sedimentation and its impact on dams and reservoirs.
Furthermore, sediment bypass tunnels as a countermeasure for small to medium sized reser-
voirs are discussed with their pros and cons. The issue of hydroabrasion is highlighted, and
the main design features to be applied for sediment bypass tunnels are given.

1  Introduction

By impounding natural watercourses, dams alter the flow regime from flowing water to a body
of standing water, which favors reservoir sedimentation. Without adequate countermeasures
ongoing sedimentation may lead to various problems such as (1) a decrease of the active
volume leading to a loss of energy production or of water available for water supply and
irrigation; (2) a decrease of the retention volume in case of flood events; (3) endangerment
of operating safety due to blockage of the outlet structures; and (4) increased abrasion of
steel hydraulics works and mechanical equipment due to increasing specific suspended load
concentrations. Besides these operational problems a lack of sediments in the downstream
river stretch may result in river bed incision. Particularly, with more severe legislation such as
the revised Swiss water protection law that has come into force in 2011, the exigencies regard-
ing ecology have increased. One of the goals is to restore the longitudinal continuity of sedi-
ments wherever possible at reasonable expense. For many smaller reservoirs, particularly in
mountainous conditions, Sediment Bypass Tunnels (SBTs) may counter these negative effects
by connecting the upstream and downstream reaches of dams and reestablishing sediment
continuity, as proven by a number of cases worldwide, particularly in Japan and Switzerland
(Auel & Boes 2011, Fukuda et al. 2012). However, due to high flow velocities and large bed
load transport rates, hydroabrasion is a frequent phenomenon present at SBT. Due to the
fact that abrasion requires continuous maintenance and causes high annual costs, adequate
countermeasures such as using High-Performance Concrete (HPC) and/or optimization of
hydraulic conditions for invert protection should be already taken into account at the design
phase (Hagmann et al. 2012).

2  Reservoir sedimentation and Countermeasures

In analogy to natural lakes, artificial reservoirs impounded by dams fill up with sediments
over time. Depending on local site conditions such as size, topography, landform, hydrology
and geology of the catchment basin, as well as size and shape of the reservoir this proc-
ess may last from a few years to several centuries. On a worldwide scale, typical sedimenta-
tion rates per country vary between a few tenths up to more than three percent (Fig.  1),

221
Figure 1. Observed sedimentation rates for various countries worldwide (after ICOLD 2009).

Figure 2.  (a) Fully-silted Koorawatha reservoir in Australia (Chanson 1998), (b) aggradation depths in
Räterichsboden reservoir from 1950 to 2001 (Bühler & Anselmetti 2003).

putting the sustainability of reservoirs into question if no adequate countermeasures are

taken. The Koorawatha reservoir in Australia, for instance, has quickly experienced consider-
able sedimentation after commissioning in 1911, so that it lost its main purpose for railway
water supply (Fig. 2a). Although such a high rate of sedimentation hardly occurs for Central
European conditions, some alpine reservoirs also show a significant aggradation process. The
aggradation depths in the Räterichsboden reservoir in the Swiss Alps amounted to 28 m after
50 years of operation (Fig. 2b).
From a hydraulics and sedimentology point of view the deposition process of bed load or
suspended load in a reservoir is described by the relationship between discharge, flow velocity
or bed shear stress and particle properties e.g. size, density and settling velocity. The aggrada-
tion pattern in a reservoir therefore depends on the kind and amount of incoming sediments
as well as the geometry and operation mode of the reservoir. Typically, due to decreasing
flow velocities and thus turbulence intensities aggraded sediments become finer from the
222
Figure  3.  (a) Reservoir sedimentation countermeasures (adapted from Sumi et al. 2004), (b) invert
abrasion at Palagnedra SBT, Switzerland (photograph by C. Auel).

upper reach of the reservoir towards the dam. A common approach distinguishes between
delta formation at the upper reach of a reservoir caused by coarse sediments (bed load) and
the aggradation of fines in the deeper water zone further downstream which is often highly
affected by density currents in the case of rather narrow and elongated reservoirs of steep
bottom slope (Schleiss et al. 2010, Boes 2011).
Reservoir sedimentation causes a number of negative impacts on dams. Firstly, when
reaching the dam it may endanger the functionality of both intake structures and bottom
outlets. Blockage of the latter must be avoided, as these constitute an important safety ele-
ment of dams. Secondly, the effective net volume available for the purpose of the reservoir, e.g.
power production or flood protection, is reduced over time due to proceeding sedimentation.
Thus aggradation of fines results in an immediate negative impact, whereas accumulation of
coarse material has a long-term negative impact on reservoirs.
To keep or restore the original reservoir volume the necessary measures are (I) prevention
of sediment input, (II) routing of incoming sediments and (III) removing aggraded sediments
a posteriori (Fig. 3a). Whereas the former have a preventive character, i.e. they impede sedi-
ments from being transported into a reservoir; the two latter methods are retroactive, as they
deal with sediments that have already been transported into the lake. Sediment bypass tun-
nels belong to the routing method, as they convey sediments around a dam into the tailwater.
SBTs are mainly operated during flood events and connect the upper and lower river reaches
and reestablish the pre-dam conditions in terms of sediment transport (sediment continuity).
In general, such measures should be taken as early as possible to maximize their efficiencies,
i.e. in the planning and design phases of dams and reservoirs. Unfortunately, despite knowl-
edge on the reservoir sedimentation process countermeasures have often been postponed or
not adequately been considered in the past, restricting the choice of efficient measures at a
later stage.

3  Sediment bypass tunnels

According to Auel & Boes (2011) a SBT consists of a guiding structure installed in the reser-
voir, an intake structure with a gate, mostly a short and steeply sloped acceleration section,
a mild sloped bypass tunnel section, and an outlet structure. Depending on the location of
223
the intake structure, i.e. whether at the head or within the reservoir, there are basically two
different types of SBTs. For type A, the inflow takes place under free surface conditions at
the delta, while for type B it is usually pressurized located below the pivot point of the aggra-
dation body. The tunnel invert has to be steep enough to avoid sediment deposition and at
the same time it should be as mild as possible to limit the flow velocities in order to prevent
invert abrasion. For existing SBTs in Japan and Switzerland, the bed slope varies between 1
and 4% (Auel & Boes 2011).
SBTs feature several advantages over other countermeasures. Firstly, they have positive
effects regarding ecological aspects, because sediment conveyance may significantly deceler-
ate or even stop river bed erosion and increase the morphological variability downstream of
a dam. Mainly sediments provided from the upstream river reach are conveyed through the
SBT since remobilization of accumulated sediments in the reservoir hardly occurs. The sedi-
ment concentration in the tailwater of the dam is thus not affected by the reservoir itself and
of natural character. Secondly, SBTs have been proven as an effective countermeasure against
reservoir sedimentation amongst others. For instance, the type A SBT of the Asahi Dam in
Japan has greatly reduced the severe aggradation in terms of accumulated sedimentation
volume after commissioning in 1998 (Fig. 4). Even during an exceptionally large flood caused
by a typhoon in 2011 the routing of sediments around the dam helped to limit the inflow of
sediments into the reservoir.
Whereas typically, the bed load deposition may be completely solved with SBT, the dep-
osition of fines depends on the design discharge of the tunnel. The higher the recurrence
interval of the SBT operation, the higher the share of the incoming suspended load that
may be conveyed through the tunnel and the smaller the amount of fines entering the res-
ervoir. The main drawback of SBT is related to economic considerations. The implemen-
tation of SBT is not only costly from an investment perspective, but also requires regular
maintenance. Due to high flow velocities with peaks in the range of 12 to 20 m/s (Auel &
Boes 2011) and high sediment transport rates, invert abrasion is generally a severe prob-
lem, requiring costly repair works and maintenance (see Fig.  3b). For this reason SBTs
should be considered as a convenient measure for small to medium-sized reservoirs with
capacity-to-inflow ratios, i.e. the ratio between the annual inflow and the total reservoir
volumes, of about 0.003 to 0.2 for the Swiss and Japanese SBTs (Sumi & Kantoush 2011).
For such reservoirs bed load aggradation and delta formation are more critical problems
than the problem of sedimentation of fines since a large amount of incoming fines stay
in suspension due to the relatively short residence time and are discharged via the outlet
works. Moreover, major tunneling costs favor their use at smaller reservoirs due to short
SBT lengths.

Figure 4. Development of reservoir sedimentation volume at Asahi reservoir, Japan, prior to and after
commissioning of an SBT in 1998 (Fukuroi 2012).

224
4  Hydroabrasion at sediment bypass tunnels

As stated above, abrasion is a serious concern in most SBTs. The extent of damages along the
wetted perimeter, i.e. mainly on the invert and the lower parts of the tunnel walls, typically
increases with increasing unit sediment load, particle size and distribution and flow velocities
i.e. shear stress as well as quartz content in the mineralogical composition of the sediments.
Hydro-abrasive damage on an invert of a hydraulic structure occurs when the flow induced
bed shear stress exceeds a critical value and hence numerous particles start impacting.
Depending on the flow conditions particles are transported in sliding, rolling or saltation
modes and cause grinding, rolling or saltating impact stress on the bed and thus wear on the
bed (Fig. 5). According to Sklar & Dietrich (2001, 2004) the governing process causing abra-
sion is saltation, whereas sliding and rolling do not cause significant wear. Therefore, for an
optimum SBT design, hydraulic conditions, particle size and distribution and hence particle
transport modes must be determined. The rolling, saltation and suspension probabilities of
different particle sizes are also important since the transport mode directly affects the particle
impact energy on the bed. They are determined using the relationship between probabili-
ties of transport modes and Shields number (Hu & Hui 1996, Ancey et al. 2002, Auel et al.
2014a). For a given particle size and flow induced shear stress, particle transport modes and
their probability as well as expected particle abrasion mechanisms can be obtained to opti-
mize the hydraulic condition and to choose the invert material type for a SBT.
Auel et al. (2014b) show that mean particle and impact velocities for rolling and saltat-
ing particles linearly increase with flow velocity independent of particle size. Large particles
possess a large particle mass, thus higher impact energy is transferred by these particles on
the surface (Auel et al. 2014a and 2014b). Consequently, a combination of large particle size
and high flow velocities results in high mean invert abrasion, which is clearly confirmed in
Figure 6. Depending on the invert material and transported sediment properties, the abra-
sion depth varies and in general typical mean abrasion depths range from microns to some
millimeter per operating hour for some Swiss and Japanese cases (Jacobs et al. 2001, Kataoka
2003, Sumi et al. 2004, and Fukuroi 2012).
The results from Runcahez SBT indicate that mean abrasion depth per hour increase with
decreasing compressive and bending tensile strengths of the invert concrete (Figs. 6a and 7).
Moreover, despite slightly larger particle size and higher design velocity compared to Runcahez
SBT, the mean abrasion depth per hour on Pfaffensprung SBT is smaller than in Runcahez.
This result reveals a strong effect of invert material properties on abrasion depth, e.g. the
higher the compressive and bending tensile strengths, the less the abrasion depth (Fig. 7).
Comparison of Runcahez SBT with Asahi SBT shows the effect of design velocity on abrasion
depth (Fig.  6b). Although the concrete compressive strength is smaller for Asahi than for
Runcahez (Fig. 7a), the high mean abrasion depth per hour in Asahi SBT in comparison to
Runcahez SBT is mostly attributed to the high design velocity despite much smaller particle
size, i.e. dm = 50 mm for Asahi vs. 225 mm for Runcahez (Fig. 6a). Note that the bed-load rates
are not known for both SBTs.
One of the important findings from the study in Pfaffensprung is the high hydroabrasion
resistance of granite compared to the implemented concrete invert, as the mean abrasion per
hour on the granite plates is five times less than on the concrete invert (Fig. 7a). This may

Figure 5.  Abrasion processes of hydraulic systems’ surfaces: (a) grinding, (b) combination of grinding
and impingement (Jacobs et al. 2001).

225
Figure 6. Observed mean abrasion depths per operating hour as functions of (a) particle diameter dm
and (b) design flow velocity vd.

Figure 7.  (a) Compressive strength determined at cubes and (b) bending tensile strength of invert con-
crete applied in various SBTs (data from studies mentioned above); symbols are the same as in Figure 6.

Figure 8.  Abrasion patterns at Pfaffensprung SBT: (a) grooves forming along joints of granite plates,
(b) undulating invert at HPC test field (steel-fibre HPC with fc > 70 N/mm2).

suggest that granite is a better choice as an invert material over HPC. However, the cost of
granite should be carefully considered at SBT design phase.
The results of the study carried out in Pfaffensprung SBT in 2013 show that the damages
typically take place in the form of grooves along the joints of basalt and granite plates (Fig. 8a),
while a wavy pattern of abrasion occurs on the HPC (Fig. 8b). In order to further reduce the
226
abrasion on the granite, this result suggests that such plates should not be implemented in
parallel to the main flow direction.
Whereas natural stone material, e.g. granite or cast basalt plates, usually have a high abra-
sion resistance against pure particle grinding action, their brittleness favors fracturing by
impinging particles in case of saltating sediments. In the latter case, either steel or cementi-
tious material such as HPC generally show a better resistance. As steel linings are often too
costly for abrasion protection of large areas such as in SBTs, HPC becomes an interesting
and economical alternative.
Based on a long-term field study performed at Runcahez SBT in Switzerland between
1995 and 2000 (Jacobs et al. 2001), the decisive material characteristics of hardened HPC
are the compressive strength (fc > 70 N/mm2 at 28 days, Fig. 7a), the bending tensile strength
(fc,bt > 10 N/mm2 at 28 days, Fig. 7b), and the fracture energy (>200 J/m2 at 28 days).

5 Design of SBT

To reduce the negative effects of hydroabrasion at SBTs, (i) an optimum hydraulic design
to limit the strong particle impact forces and (ii) a selection of sustainable and optimum
abrasion-resistant invert lining material are recommended measures. Aspect (i) demands for
the following, among others:
• A tunnel cross section with plane invert geometry should be chosen, i.e. archway and horse-
shoe profiles with horizontal bed rather than circular ones, to avoid stress concentrations.
• Whenever possible, bends in plan view should be avoided to reduce shock waves and sec-
ondary currents, which cause locally high specific sediment transport rates. For instance,
in Solis SBT the sediment transport at the tunnel outlet is concentrated on the orographic
right side as a result of a bend further upstream (Fig. 9a).
• Keep the bed slope as mild as possible without endangering sediment aggradation.
As to aspect (ii) the following should be accounted for, among others:
• If most of the particles are transported in rolling or sliding motion with only minor
saltation, abrasion processes are expected to be mainly grinding and only weakly
impinging. Hence using natural stones such as granite and cast basalt as invert lining
material is a good solution. As the joints should not be parallel to the flow, the use
of hexagonal plates is recommended (Fig. 9b). The plates should be embedded into a
special mortar.
• If saltation is expected to be the main particle transport mode and/or the sediment is
rather coarse with high flow velocity, HPC with compressive strength above 70 N/mm2
(i.e. C70/85 and higher) and a bending tensile strength above 10 N/mm2 is preferable.
Concrete curing is critical and should be carefully performed. For more details, see
Jacobs et al. (2001).

Figure 9.  (a) Measured sediment transport using 8 geophones for the 2013 flood event at Solis SBT;
(b) schematic plan view of a thin pavement of hexagonal natural stone plates (Jacobs et al. 2001).

227
 6  Conclusions

Reservoir sedimentation is a serious worldwide problem threatening the sustainability of res-

ervoirs and negatively impacting dam safety. SBTs are effective countermeasures for small to
medium-sized reservoirs provided that hydroabrasion is accounted for from the very begin-
ning of the planning stage by optimizing the hydraulic design and by applying adequate
invert material. Depending on the transport mode of the sediment which is determined by
the hydraulic characteristics of the tunnel flow, cast basalt and granite plates or HPC and
steel linings are recommended.

REFERENCES

Ancey, C., Bigillon, F., Frey, P., Lanier, J. & Ducret, R. 2002. Saltating motion of a bead in a rapid water
stream. Physical Review E 66(3).
Auel, C., Albayrak, I. & Boes, R.M. 2014a. Particle motion in supercritical open channel flows. Journal
of Hydraulic Engineering: submitted.
Auel, C., Albayrak, I. & Boes, R.M. 2014b. Bedload particle velocity in supercritical open channel flows.
Proc. Intl. River Flow Conference, EPF Lausanne, Switzerland: submitted.
Auel, C. & Boes, R.M. 2011: Sediment bypass tunnel design—review and outlook. Proc. ICOLD Sym-
posium “Dams under changing challenges” (A.J. Schleiss & R.M. Boes, eds.), 79th Annual Meeting,
Lucerne. Taylor & Francis, London: 403–412.
Boes, R. 2011. Nachhaltigkeit von Talsperren angesichts der Stauraumverlandung (‘Sustainability of
dams in view of reservoir sedimentation’). Mitteilung 164, Lehrstuhl und Institut für Wasserbau und
Wasserwirtschaft (H. Schüttrumpf, ed.), RWTH Aachen, Germany: 161–174 [in German].
Bühler, R. & Anselmetti, F. 2003. Ablagerungen in den Grimsel-Stauseen. Teil B: Räterichsbodensee
(‘Deposition in the Grimsel reservoirs. Part B: Räterichsboden reservoir’). Report Limnogeology lab
of the Geological Institute, ETH Zurich, Switzerland [in German].
Chanson, H. 1998. Extreme reservoir sedimentation in Australia: a Review. Intl Jl. of Sediment Research
13(3): 55–63 (ISSN 1001-6279).
Fukuda, T., Yamashita, K., Osada, K. & Fukuoka, S. 2012. Study on Flushing Mechanism of Dam
Reservoir Sedimentation and Recovery of Riffle-Pool in Downstream Reach by a Flushing Bypass
Tunnel. Proc. Intl. Symposium on Dams for a changing world, Kyoto, Japan.
Fukuroi, H. 2012. Damage from Typhoon Talas to Civil Engineering Structures for Hydropower and
the Effect of the Sediment Bypass System at Asahi Dam. Proc. Int. Symposium on Dams for a chang-
ing World—Need for Knowledge Transfer across the Generations and the World. Kyoto, Japan.
Hagmann, M., Albayrak, I. & Boes, R.M. 2012. Reduktion der Hydroabrasion bei Sediment­
umleitstollen—In-situ-Versuche zur Optimierung der Abrasionsresistenz (‘Reduction of hydroa-
brasion in sediment bypass tunnels—in-situ experiments to optimize abrasion resistance‘). Proc.
Wasserbausymposium, TU Graz (G. Zenz, ed.), A12: 91–98 [in German].
Hu, C. & Hui, Y. 1996. Bed-load transport. I: Mechanical characteristics. Journal of Hydraulic
­Engineering 122(5), 245–254.
ICOLD 2009. Sedimentation and Sustainable Use of Reservoirs and River Systems. Draft Bulletin 147,
Paris, France.
Jacobs, F., Winkler, K., Hunkeler, F. & Volkart, P. 2001. Betonabrasion im Wasserbau (‘Concrete abra-
sion at hydraulic structures’). VAW-Mitteilung 168 (H.-E. Minor, ed.), VAW, ETH Zurich, ­Switzerland
[in German].
Kataoka, K. 2003. Sedimentation Management at Asahi Dam. Proc. Third World Water Forum, Siga,
Japan, 197–207.
Schleiss, A., De Cesare, G. & Jenzer Althaus, J. 2010. Verlandung der Stauseen gefährdet die nachhaltige
Nutzung der Wasserkraft (‘Reservoir sedimentation threatens the sustainable use of hydropower’),
Wasser, Energie, Luft, 102(1): 31–40 [in German].
Sklar, L.S. & Dietrich, W.E. 2001. Sediment and rock strength controls on river incision into bedrock.
Geology 29(12), 1087–1090.
Sklar, L.S. & Dietrich, W.E. 2004. A mechanistic model for river incision into bedrock by saltating bed
load. Water Resources Research 40(W06301), 21p.
Sumi, T. & Kantoush, S.A. 2011. Comprehensive Sediment Management Strategies in Japan: Sediment
bypass tunnels. Proc. 34th IAHR World Congress. Brisbane, Australia, 1803–1810.
Sumi, T., Okano, M. & Takata, Y. 2004. Reservoir sedimentation management with bypass tunnels in
Japan. Proc. 9th International Symposium on River Sedimentation. Yichang, China, 1036–1043.

228

View publication stats